Section 8.5 Taylor Polynomials and Taylor Series
ΒΆMotivating Questions
What is a Taylor polynomial? For what purposes are Taylor polynomials used?
What is a Taylor series?
How do we determine the accuracy when we use a Taylor polynomial to approximate a function?
Preview Activity 8.5.1.
Preview Activity 8.3.1 showed how we can approximate the number ee using linear, quadratic, and other polynomial functions; we then used similar ideas in Preview Activity 8.4.1 to approximate ln(2).ln(2). In this activity, we review and extend the process to find the βbestβ quadratic approximation to the exponential function exex around the origin. Let f(x)=exf(x)=ex throughout this activity.
Find a formula for P1(x),P1(x), the linearization of f(x)f(x) at x=0.x=0. (We label this linearization P1P1 because it is a first degree polynomial approximation.) Recall that P1(x)P1(x) is a good approximation to f(x)f(x) for values of xx close to 0.0. Plot ff and P1P1 near x=0x=0 to illustrate this fact.
-
Since f(x)=exf(x)=ex is not linear, the linear approximation eventually is not a very good one. To obtain better approximations, we want to develop a different approximation that βbendsβ to make it more closely fit the graph of ff near x=0.x=0. To do so, we add a quadratic term to P1(x).P1(x). In other words, we let
P2(x)=P1(x)+c2x2P2(x)=P1(x)+c2x2for some real number c2.c2. We need to determine the value of c2c2 that makes the graph of P2(x)P2(x) best fit the graph of f(x)f(x) near x=0.x=0.
Remember that P1(x)P1(x) was a good linear approximation to f(x)f(x) near 0;0; this is because P1(0)=f(0)P1(0)=f(0) and Pβ²1(0)=fβ²(0).Pβ²1(0)=fβ²(0). It is therefore reasonable to seek a value of c2c2 so that
P2(0)=f(0),Pβ²2(0)=fβ²(0),and Pβ³2(0)=fβ³(0).P2(0)=f(0),Pβ²2(0)=fβ²(0),and Pβ²β²2(0)=fβ²β²(0).Remember, we are letting P2(x)=P1(x)+c2x2.
Calculate P2(0) to show that P2(0)=f(0).
Calculate Pβ²2(0) to show that Pβ²2(0)=fβ²(0).
Calculate Pβ³2(x). Then find a value for c2 so that Pβ³2(0)=fβ³(0).
Explain why the condition Pβ³2(0)=fβ³(0) will put an appropriate βbendβ in the graph of P2 to make P2 fit the graph of f around x=0.
-
We know that
\begin{equation*} P_1(x) = f(0) + f'(0)x = 1+x\text{.} \end{equation*}Since \(P_1(0) = f(0) = 1\) and \(P'_1(0) = f'(0) = 1\text{,}\) the graphs of \(P_1\) and \(f\) agree at \(x=a\) and have the same slope at \(x=0\) (which means they go in the same direction at \(x=0\)). This is why \(P_1(x)\) is a good approximation to \(f(x)\) for values of \(x\) close to \(0\text{.}\)
-
Since
\begin{equation*} P_2(x) = P_1(x) + c_2(x)^2 = f(0) + f'(0)x + c_2x^2 \end{equation*}we have that
\begin{equation*} P_2(0) = 1 = f(0) \end{equation*}as desired.
A simple calculation shows \(P'_2(x) = P'1(x) + 2c_2x\text{.}\) So \(P'_2(0) = P'_1(0) = 1 = f'(0)\) as desired.
A simple calculation shows \(P''_2(x) = 2c_2\text{.}\) So \(P''_2(0) = 2c_2\text{.}\) To have \(P''_2(0) = f''(0)\) we must have \(2c_2 = f''(0)\) or \(c_2 = \frac{f''(0)}{2} = \frac{1}{2}\text{.}\)
The second derivative of a function tells us the concavity of the function. Concavity measures how the slopes of the tangent lines to the graph of the function are changing. This tells us how much bend there is in the graph. So if \(P''_2(0) = f''(0)\text{,}\) then \(P_2\) will have the same bend in it at \(x=0\) as \(f\) does. This will make the graph of \(P_2\) mold to the graph of \(f\) around \(x=0\text{.}\)
-
Subsection 8.5.1 Taylor Polynomials
Preview Activity 8.5.1 illustrates the first steps in the process of approximating functions with polynomials. Using this process we can approximate trigonometric, exponential, logarithmic, and other nonpolynomial functions as closely as we like (for certain values of x) with polynomials. This is extraordinarily useful in that it allows us to calculate values of these functions to whatever precision we like using only the operations of addition, subtraction, multiplication, and division, which can be easily programmed in a computer. We next extend the approach in Preview Activity 8.5.1 to arbitrary functions at arbitrary points. Let f be a function that has as many derivatives as we need at a point x=a. Recall that P1(x) is the tangent line to f at (a,f(a)) and is given by the formulaTaylor Polynomials.
The nth order Taylor polynomial of f centered at x=a is given by
This degree n polynomial approximates f(x) near x=a and has the property that P(k)n(a)=f(k)(a) for k=0,1,β¦,n.
Example 8.5.1.
Determine the third order Taylor polynomial for \(f(x) = e^x\text{,}\) as well as the general \(n\)th order Taylor polynomial for \(f\) centered at \(x=0\text{.}\)
We know that \(f'(x) = e^x\) and so \(f''(x) = e^x\) and \(f'''(x) = e^x\text{.}\) Thus,
So the third order Taylor polynomial of \(f(x) = e^x\) centered at \(x=0\) is
In general, for the exponential function \(f\) we have \(f^{(k)}(x) = e^x\) for every positive integer \(k\text{.}\) Thus, the \(k\)th term in the \(n\)th order Taylor polynomial for \(f(x)\) centered at \(x=0\) is
Therefore, the \(n\)th order Taylor polynomial for \(f(x) = e^x\) centered at \(x=0\) is
Activity 8.5.2.
We have just seen that the nth order Taylor polynomial centered at a=0 for the exponential function ex is
In this activity, we determine small order Taylor polynomials for several other familiar functions, and look for general patterns.
-
Let f(x)=11βx.
Calculate the first four derivatives of f(x) at x=0. Then find the fourth order Taylor polynomial P4(x) for 11βx centered at 0.
Based on your results from part (i), determine a general formula for f(k)(0).
-
Let f(x)=cos(x).
Calculate the first four derivatives of f(x) at x=0. Then find the fourth order Taylor polynomial P4(x) for cos(x) centered at 0.
Based on your results from part (i), find a general formula for f(k)(0). (Think about how k being even or odd affects the value of the kth derivative.)
-
Let f(x)=sin(x).
Calculate the first four derivatives of f(x) at x=0. Then find the fourth order Taylor polynomial P4(x) for sin(x) centered at 0.
Based on your results from part (i), find a general formula for f(k)(0). (Think about how k being even or odd affects the value of the kth derivative.)
Small hints for each of the prompts above.
\(f^{k}(0) = 0 \text{ if } k \text{ is odd, and } f^{2k}(0) = (-1)^k \text{.}\)
\(P_n(x) = 1 - \frac{x^2}{2} + \frac{x^4}{4!} - \frac{x^6}{6!} + \cdots + (-1)^{n/2}\frac{x^n}{n!}\) if \(n\) is even and \(P_n(x) = 1 - \frac{x^2}{2} + \frac{x^4}{4!} - \frac{x^6}{6!} + \cdots + (-1)^{(n-1)/2}\frac{x^(n-1)}{(n-1)!}\) if \(n\) is odd.
\(f^{k}(0) = 0 \text{ if } k \text{ is even } \ \ \ \text{ and } \ \ \ f^{2k+1}(0) = (-1)^k \text{.}\)
\(P_n(x) = x - \frac{x^3}{3!} + \frac{x^5}{5!} - \frac{x^7}{7!} + \cdots + (-1)^{(n-1)/2}\frac{x^n}{n!}\) if \(n\) is odd and \(P_n(x) = x - \frac{x^3}{3!} + \frac{x^5}{5!} - \frac{x^7}{7!} + \cdots + (-1)^{n/2+1}\frac{x^{n-1}}{(n-1)!}\) if \(n\) is even.
\(f^{(k)}(0) = k! \text{.}\)
- \begin{equation*} P_n(x) = \sum_{k=0}^n x^k\text{.} \end{equation*}
-
The first four derivatives of \(f(x)\) at \(x=0\) are
\begin{align*} f(x) \amp = \cos(x) \amp f(0) \amp = 1\\ f'(x) \amp = -\sin(x) \amp f'(0) \amp = 0\\ f''(x) \amp = -\cos(x) \amp f''(0) \amp = -1\\ f^{(3)}(x) \amp = \sin(x) \amp f^{(3)}(0) \amp = 0\\ f^{(4)}(x) \amp = \cos(x) \amp f^{(4)}(0) \amp = 1\text{.} \end{align*}It appears that the odd derivatives of \(f(x)\) are all plus or minus \(\sin(x)\) and so have values of 0 at \(x=0\) and the even derivatives are \(\pm \cos(x)\) and have alternating values of 1 and \(-1\) at \(x-0\text{.}\) Since the even numbers can be represented in the form \(2k\) where \(k\) is an integer we have
\begin{equation*} f^{k}(0) = 0 \text{ if } k \text{ is odd, and } f^{2k}(0) = (-1)^k\text{.} \end{equation*} -
Based on the previous part of this problem the \(n\)th order Taylor polynomial for \(\cos(x)\) is
\begin{equation*} 1 - \frac{x^2}{2} + \frac{x^4}{4!} - \frac{x^6}{6!} + \cdots + (-1)^{n/2}\frac{x^n}{n!} \end{equation*}if \(n\) is even and
\begin{equation*} 1 - \frac{x^2}{2} + \frac{x^4}{4!} - \frac{x^6}{6!} + \cdots + (-1)^{(n-1)/2}\frac{x^(n-1)}{(n-1)!} \end{equation*}if \(n\) is odd.
-
-
The first four derivatives of \(f(x)\) at \(x=0\) are
\begin{align*} f(x) \amp = \sin(x) \amp f(0) \amp = 0\\ f'(x) \amp = \cos(x) \amp f'(0) \amp = 1\\ f''(x) \amp = -\sin(x) \amp f''(0) \amp = 0\\ f^{(3)}(x) \amp = -\cos(x) \amp f^{(3)}(0) \amp = -1\\ f^{(4)}(x) \amp = \sin(x) \amp f^{(4)}(0) \amp = 0\text{.} \end{align*}It appears that the even derivatives of \(f(x)\) are all plus or minus \(\sin(x)\) and so have values of 0 at \(x=0\) and the odd derivatives are \(\pm \cos(x)\) and have alternating values of 1 and \(-1\) at \(x-0\text{.}\) Since the odd numbers can be represented in the form \(2k+1\) where \(k\) is an integer we have
\begin{equation*} f^{k}(0) = 0 \text{ if } k \text{ is even } \ \ \ \text{ and } \ \ \ f^{2k+1}(0) = (-1)^k\text{.} \end{equation*} -
Based on the previous part of this problem the \(n\)th order Taylor polynomial for \(\sin(x)\) is
\begin{equation*} x - \frac{x^3}{3!} + \frac{x^5}{5!} - \frac{x^7}{7!} + \cdots + (-1)^{(n-1)/2}\frac{x^n}{n!} \end{equation*}if \(n\) is odd and
\begin{equation*} x - \frac{x^3}{3!} + \frac{x^5}{5!} - \frac{x^7}{7!} + \cdots + (-1)^{n/2+1}\frac{x^{n-1}}{(n-1)!} \end{equation*}if \(n\) is even.
-
-
The first four derivatives of \(f(x)\) at \(x=0\) are
\begin{align*} f(x) \amp = \frac{1}{1-x} \amp f(0) \amp = 1\\ f'(x) \amp = \frac{1}{(1-x)^2} \amp f'(0) \amp = 1\\ f''(x) \amp = \frac{2}{(1-x)^3} \amp f''(0) \amp = 2\\ f^{(3)}(x) \amp = \frac{3!}{(1-x)^4} \amp f^{(3)}(0) \amp = 3!\\ f^{(4)}(x) \amp = \frac{4!}{(1-x)^5} \amp f^{(4)}(0) \amp = 4!\text{.} \end{align*}It appears that the pattern is
\begin{equation*} f^{(k)}(0) = k!\text{.} \end{equation*} -
The \(n\)th order Taylor polynomial for \(f\) at \(x=0\) is
\begin{equation*} \sum_{k=0}^n \frac{f^{(k)}}{k!} x^k = \sum_{k=0}^n \frac{k!}{k!} x^k = \sum_{k=0}^n x^k\text{.} \end{equation*}This makes sense since \(f(x)\) is the sum of the geometric series with ratio \(x\text{,}\) so the \(n\)th order Taylor polynomial should just be the \(n\)th partial sum of this geometric series.
-
Subsection 8.5.2 Taylor Series
In Activity 8.5.2 we saw that the fourth order Taylor polynomial P4(x) for sin(x) centered at 0 isDefinition 8.5.3.
Let f be a function all of whose derivatives exist at x=a. The Taylor series for f centered at x=a is the series Tf(x) defined by
Activity 8.5.3.
In Activity 8.5.2 we determined small order Taylor polynomials for a few familiar functions, and also found general patterns in the derivatives evaluated at 0. Use that information to write the Taylor series centered at 0 for the following functions.
f(x)=11βx
f(x)=cos(x) (You will need to carefully consider how to indicate that many of the coefficients are 0. Think about a general way to represent an even integer.)
f(x)=sin(x) (You will need to carefully consider how to indicate that many of the coefficients are 0. Think about a general way to represent an odd integer.)
Determine the n order Taylor polynomial for f(x)=11βx centered at x=0.
Small hints for each of the prompts above.
\(P(x) = 1 + x + x^2 + x^3 + \cdots + x^n + \cdots\)
\(P(x) = 1 - \frac{1}{2!}x^2 + \frac{1}{4!}x^4 - \cdots + (-1)^{n}\frac{1}{(2n)!}x^{2n} + \cdots \text{.}\)
\(P(x) = x - \frac{1}{3!}x^3 + \frac{1}{5!}x^5 - \cdots + (-1)^{n}\frac{1}{(2n+1)!}x^{2n+1} + \cdots \text{.}\)
\(P_n(x) = 1 + x + x^2 + x^3 + \cdots + x^n\)
-
For \(f(x) = \frac{1}{1-x}\text{,}\) its Taylor series is
\begin{equation*} P(x) = 1 + x + x^2 + x^3 + \cdots + x^n + \cdots \end{equation*} -
For \(f(x) = \cos(x)\text{,}\) its Taylor series is
\begin{equation*} P(x) = 1 - \frac{1}{2!}x^2 + \frac{1}{4!}x^4 - \cdots + (-1)^{n}\frac{1}{(2n)!}x^{2n} + \cdots\text{.} \end{equation*} -
For \(f(x) = \sin(x)\text{,}\) its Taylor series is
\begin{equation*} P(x) = x - \frac{1}{3!}x^3 + \frac{1}{5!}x^5 - \cdots + (-1)^{n}\frac{1}{(2n+1)!}x^{2n+1} + \cdots\text{.} \end{equation*} -
For \(f(x) = \frac{1}{1-x}\text{,}\)
\begin{equation*} P_n(x) = 1 + x + x^2 + x^3 + \cdots + x^n \end{equation*}
Activity 8.5.4.
Plot the graphs of several of the Taylor polynomials centered at 0 (of order at least 5) for ex and convince yourself that these Taylor polynomials converge to ex for every value of x.
Draw the graphs of several of the Taylor polynomials centered at 0 (of order at least 6) for cos(x) and convince yourself that these Taylor polynomials converge to cos(x) for every value of x. Write the Taylor series centered at 0 for cos(x).
Draw the graphs of several of the Taylor polynomials centered at 0 for 11βx. Based on your graphs, for what values of x do these Taylor polynomials appear to converge to 11βx? How is this situation different from what we observe with ex and cos(x)? In addition, write the Taylor series centered at 0 for 11βx.
Small hints for each of the prompts above.
It appears that as we increase the order of the Taylor polynomials, they fit the graph of \(f\) better and better over larger intervals.
It appears that as we increase the order of the Taylor polynomials, they fit the graph of \(f\) better and better over larger intervals.
The Taylor polynomials converge to \(\frac{1}{1-x}\) only on the interval \((-1,1)\text{.}\)
-
The graphs of the 10th (magenta), 20th (blue), and 30th (green) Taylor polynomials centered at \(0\) for \(e^x\) are shown below along with the graph of \(f(x)\) in red:
It appears that as we increase the order of the Taylor polynomials, they fit the graph of \(f\) better and better over larger intervals. So it looks like the Taylor polynomials converge to \(e^x\) for every value of \(x\text{.}\)
-
The graphs of the 10th (magenta), 20th (blue), and 30th (green) Taylor polynomials centered at \(0\) for \(\cos(x)\) are shown below along with the graph of \(f(x)\) in red:
It appears that as we increase the order of the Taylor polynomials, they fit the graph of \(f\) better and better over larger intervals. So it looks like the Taylor polynomials converge to \(\cos(x)\) for every value of \(x\text{.}\) Based on the \(n\)th order Taylor polynomials we found earlier for \(\cos(x)\text{,}\) the Taylor series for \(f(x)\) centered at \(0\) is
\begin{equation*} \sum_{k=0}^{\infty} \frac{x^{2k}}{(2k)!}\text{.} \end{equation*} -
The graphs of the 10th (magenta), 20th (blue), and 30th (green) Taylor polynomials centered at \(0\) for \(\frac{1}{1-x}\) are shown below along with the graph of \(f(x)\) in red:
It appears that as we increase the order of the Taylor polynomials, they only fit the graph of \(f\) better and better over the interval \((-1,1)\) and appear to diverge outside that interval. So it looks like the Taylor polynomials converge to \(\frac{1}{1-x}\) only on the interval \((-1,1)\text{.}\)
Based on the \(n\)th order Taylor polynomials we found earlier for \(\frac{1}{1-x}\text{,}\) the Taylor series for \(f(x)\) centered at \(0\) is
\begin{equation*} \sum_{k=0}^{\infty} x^k\text{.} \end{equation*}
Subsection 8.5.3 The Interval of Convergence of a Taylor Series
In the previous section (in Figure 8.5.2 and Activity 8.5.4) we observed that the Taylor polynomials centered at 0 for ex, cos(x), and sin(x) converged to these functions for all values of x in their domain, but that the Taylor polynomials centered at 0 for 11βx converge to 11βx on the interval (β1,1) and diverge for all other values of x. So the Taylor series for a function f(x) does not need to converge for all values of x in the domain of f. Our observations suggest two natural questions: can we determine the values of x for which a given Taylor series converges? And does the Taylor series for a function f actually converge to f(x)?Example 8.5.4.
Graphical evidence suggests that the Taylor series centered at \(0\) for \(e^x\) converges for all values of \(x\text{.}\) To verify this, use the Ratio Test to determine all values of \(x\) for which the Taylor series
converges absolutely.
Recall that the Ratio Test applies only to series of nonnegative terms. In this example, the variable \(x\) may have negative values. But we are interested in absolute convergence, so we apply the Ratio Test to the series
Now, observe that
for any value of \(x\text{.}\) So the Taylor series (8.5.4) converges absolutely for every value of \(x\text{,}\) and thus converges for every value of \(x\text{.}\)
If L=0, then the Taylor series converges on (ββ,β).
If L is infinite, then the Taylor series converges only at x=a.
-
If L is finite and nonzero, then the Taylor series converges absolutely for all x that satisfy
|xβa|β L<1or for all x such that
|xβa|<1L,which is the interval
(aβ1L,a+1L).Because the Ratio Test is inconclusive when the |xβa|β L=1, the endpoints aΒ±1L have to be checked separately.
Activity 8.5.5.
Use the Ratio Test to explicitly determine the interval of convergence of the Taylor series for f(x)=11βx centered at x=0.
Use the Ratio Test to explicitly determine the interval of convergence of the Taylor series for f(x)=cos(x) centered at x=0.
Use the Ratio Test to explicitly determine the interval of convergence of the Taylor series for f(x)=sin(x) centered at x=0.
Small hints for each of the prompts above.
\((-\infty, \infty)\text{.}\)
\((-\infty, \infty)\text{.}\)
The interval \((-1,1)\text{.}\)
-
Using the Ratio Test with the \(k\)th term \(\frac{|x|^{2k}}{(2k)!}\) we get
\begin{align*} \lim_{k \to \infty} \frac{ \frac{|x|^{2(k+1)}}{(2(k+1))!} }{ \frac{|x|^{2k}}{(2k)!} } \amp = \lim_{k \to \infty} \frac{|x|^{2(k+1)}(2k)!}{|x|^{2k}(2(k+1))!}\\ \amp = \lim_{k \to \infty} \frac{|x|^{2}}{(2k+2)(2k+1)}\\ \amp = 0\text{.} \end{align*}So the interval of convergence of the Taylor series for \(f(x) = \cos(x)\) centered at \(x=0\) is \((-\infty, \infty)\text{.}\)
-
Using the Ratio Test with the \(k\)th term \(\frac{|x|^{2k+1}}{(2k+1)!}\) we get
\begin{align*} \lim_{k \to \infty} \frac{ \frac{|x|^{2(k+1)+1}}{(2(k+1)+1)!} }{ \frac{|x|^{2k+1}}{(2k+1)!} } \amp = \lim_{k \to \infty} \frac{|x|^{2(k+1)+1}(2k+1)!}{|x|^{2k+1}(2(k+1)+1)!}\\ \amp = \lim_{k \to \infty} \frac{|x|^{2}}{(2k+3)(2k+2)}\\ \amp = 0\text{.} \end{align*}So the interval of convergence of the Taylor series for \(f(x) = \sin(x)\) centered at \(x=0\) is \((-\infty, \infty)\text{.}\)
-
Using the Ratio Test with the \(k\)th term \(|x|^{k}\) we get
\begin{equation*} \lim_{k \to \infty} \frac{ |x|^{k+1} }{ |x|^{k} } = \lim_{k \to \infty} |x| = |x|\text{,} \end{equation*}So the series \(\sum_{k=0}^{\infty} x^k\) converges absolutely when \(|x| \lt 1\) or for \(-1 \lt x \lt 1\) and diverges when \(|x| \gt 1\text{.}\) Since the Ratio Test doesn't tell us what happens when \(x=1\text{,}\) we need to check the endpoints separately.
When \(x=1\) we have the series \(\sum_{k=0}^{\infty} 1\) which diverges since \(\lim_{k \to \infty} 1 \neq 0\text{.}\)
When \(x=-1\) we have the series \(\sum_{k=0}^{\infty} (-1)^k\) which diverges since \(\lim_{k \to \infty} (-1)^k\) does not exist.
Therefore, the interval of convergence of the Taylor series for \(f(x) = \frac{1}{1-x}\) centered at \(x=0\) is \((-1,1)\text{.}\)
Subsection 8.5.4 Error Approximations for Taylor Polynomials
We now know how to find Taylor polynomials for functions such as sin(x), and how to determine the interval of convergence of the corresponding Taylor series. We next develop an error bound that will tell us how well an nth order Taylor polynomial Pn(x) approximates its generating function f(x). This error bound will also allow us to determine whether a Taylor series on its interval of convergence actually equals the function f from which the Taylor series is derived. Finally, we will be able to use the error bound to determine the order of the Taylor polynomial Pn(x) that we will ensure that Pn(x) approximates f(x) to the desired degree of accuracy. For this argument, we assume throughout that we center our approximations at 0 (but a similar argument holds for approximations centered at a). We define the exact error, En(x), that results from approximating f(x) with Pn(x) byThe Lagrange Error Bound for Pn(x).
Let f be a continuous function with n+1 continuous derivatives. Suppose that M is a positive real number such that |f(n+1)(x)|β€M on the interval [a,c]. If Pn(x) is the nth order Taylor polynomial for f(x) centered at x=a, then
Example 8.5.5.
Determine how well the 10th order Taylor polynomial \(P_{10}(x)\) for \(\sin(x)\text{,}\) centered at \(0\text{,}\) approximates \(\sin(2)\text{.}\)
To answer this question we use \(f(x) = \sin(x)\text{,}\) \(c = 2\text{,}\) \(a=0\text{,}\) and \(n = 10\) in the Lagrange error bound formula. We also need to find an appropriate value for \(M\text{.}\) Note that the derivatives of \(f(x) = \sin(x)\) are all equal to \(\pm \sin(x)\) or \(\pm \cos(x)\text{.}\) Thus,
for any \(n\) and \(x\text{.}\) Therefore, we can choose \(M\) to be \(1\text{.}\) Then
So \(P_{10}(2)\) approximates \(\sin(2)\) to within at most \(0.00005130671797\text{.}\) A computer algebra system tells us that
with an actual difference of about \(0.0000500159\text{.}\)
Activity 8.5.6.
Let Pn(x) be the nth order Taylor polynomial for sin(x) centered at x=0. Determine how large we need to choose n so that Pn(2) approximates sin(2) to 20 decimal places.
Small hints for each of the prompts above.
\(n \ge 27\text{.}\)
In this example, if we can find a value of \(n\) so that
then we will have
Again we use \(f(x) = \sin(x)\text{,}\) \(c = 2\text{,}\) \(a=0\text{,}\) and \(M = 1\) from the previous example. So we need to find \(n\) to make
There is no good way to solve equations involving factorials, so we simply use trial and error, evaluating \(\frac{2^{n+1}}{(n+1)!}\) at different values of \(n\) until we get one we need.
\(n\) | \(\frac{2^{n+1}}{(n+1)!}\) |
\(10\) | \(5.130671797 \times 10^{-5}\) |
\(20\) | \(4.104743250 \times 10^{-14}\) |
\(25\) | \(1.664028884 \times 10^{-19}\) |
\(26\) | \(1.232613988 \times 10^{-20}\) |
\(27\) | \(8.804385630 \times 10^{-22}\) |
So we need to use an \(n\) of at least 27 to ensure accuracy to 20 decimal places.
A computer algebra system gives
and we can see that these agree to 20 places.
Example 8.5.6.
Show that the Taylor series for \(\sin(x)\) actually converges to \(\sin(x)\) for all \(x\text{.}\)
Recall from the previous example that since \(f(x) = \sin(x)\text{,}\) we know
for any \(n\) and \(x\text{.}\) This allows us to choose \(M = 1\) in the Lagrange error bound formula. Thus,
for every \(x\text{.}\)
We showed in earlier work that the Taylor series \(\sum_{k=0}^{\infty} \frac{x^k}{k!}\) converges for every value of \(x\text{.}\) Because the terms of any convergent series must approach zero, it follows that
for every value of \(x\text{.}\) Thus, taking the limit as \(n \to \infty\) in the inequality (8.5.6), it follows that
As a result, we can now write
for every real number \(x\text{.}\)
Activity 8.5.7.
Show that the Taylor series centered at 0 for cos(x) converges to cos(x) for every real number x.
-
Next we consider the Taylor series for ex.
Show that the Taylor series centered at 0 for ex converges to ex for every nonnegative value of x.
Show that the Taylor series centered at 0 for ex converges to ex for every negative value of x.
Explain why the Taylor series centered at 0 for ex converges to ex for every real number x. Recall that we earlier showed that the Taylor series centered at 0 for ex converges for all x, and we have now completed the argument that the Taylor series for ex actually converges to ex for all x.
Let Pn(x) be the nth order Taylor polynomial for ex centered at 0. Find a value of n so that Pn(5) approximates e5 correct to 8 decimal places.
Small hints for each of the prompts above.
Compare Example 8.5.6.
Use the fact that that \(|f^{(n)}(x)| \le e^c\) on the interval \([0,c]\) for any fixed positive value of \(c\text{.}\)
Repeat the argument in (a) but replace \(e^c\) with \(1\text{,}\) and everything else holds in the same way.
Combine the results of (a) and (b)
\(n = 28\text{.}\)
Compare Example 8.5.6.
-
Let \(x \ge 0\text{.}\) Since \(f(x) = e^x\text{,}\) \(f^{(n)}(x) = e^x\) for every natural number \(n\text{.}\) Since \(e^x\) is an increasing function, we know that \(|f^{(n)}(x)| \le e^c\) on the interval \([0,c]\) for any fixed positive value of \(c\text{.}\) Thus, by the Lagrange error formula, we can say that
\begin{equation*} |P_n(x) - e^x| \le e^c \frac{x^{n+1}}{(n+1)!}\text{.} \end{equation*}Since the series \(\sum \frac{x^{n}}{n!}\) converges for every \(x\text{,}\) \(\frac{x^{n}}{n!} \to 0\) as \(x \to \infty\text{,}\) and thus \(\frac{x^{n+1}}{(n+1)!} \to 0\) as \(n \to \infty\) for every \(x\) in \([0,c]\text{.}\) Further, since \(e^c\) is a constant independent of \(n\text{,}\) \(e^c \frac{x^{n+1}}{(n+1)!} \to 0\) as well. Thus,
\begin{equation*} \lim_{n \to \infty} |P_n(x) - e^x| = 0\text{,} \end{equation*}as desired.
When \(x \lt 0\text{,}\) we know \(e^x \lt 1\text{.}\) Thus, we can repeat our argument in (a) but replace \(e^c\) with \(1\text{,}\) and everything else holds in the same way.
Because we have shown that the Taylor series for \(e^x\) converges to \(e^x\) for both every nonnegative \(x\)-value and for every negative \(x\)-value, it follows that we have convergence for every value of \(x\text{.}\)
-
-
Since \(e^x\) is increasing on \([0,5]\) we know that \(e^x \lt e^5\) on \([0,5]\text{.}\) Now \(e^5 \lt 243\text{,}\) so
\begin{equation*} \left|P_n(5) - e^5\right| \leq 243\frac{|5|^{n+1}}{(n+1)!}\text{.} \end{equation*}We want a value of \(n\)that makes this error term less than \(10^{-8}\text{.}\) Testing various values of \(n\) gives
\begin{equation*} 243\frac{|5|^{28+1}}{(28+1)!} \approx 5.119146745 \times 10^{-9} \end{equation*}so we can choose \(n = 28\text{.}\) A computer algebra system shows that \(P_{28}(5) \approx 148.413159102551\) while \(e^5 \approx 148.413159102577\) and we can see that these two approximations agree to 8 decimal places.
Subsection 8.5.5 Summary
-
We can use Taylor polynomials to approximate functions. This allows us to approximate values of functions using only addition, subtraction, multiplication, and division of real numbers. The nth order Taylor polynomial centered at x=a of a function f is
Pn(x)=(f(a)+fβ²(a)(xβa)+fβ³(a)2!(xβa)2+β―+f(n)(a)n!(xβa)n=(nβk=0f(k)(a)k!(xβa)k. -
The Taylor series centered at x=a for a function f is
ββk=0f(k)(a)k!(xβa)k.The nth order Taylor polynomial centered at a for f is the nth partial sum of its Taylor series centered at a. So the nth order Taylor polynomial for a function f is an approximation to f on the interval where the Taylor series converges; for the values of x for which the Taylor series converges to f we write
f(x)=ββk=0f(k)(a)k!(xβa)k. -
The Lagrange Error Bound shows us how to determine the accuracy in using a Taylor polynomial to approximate a function. More specifically, if Pn(x) is the nth order Taylor polynomial for f centered at x=a and if M is an upper bound for |f(n+1)(x)| on the interval [a,c], then
|Pn(c)βf(c)|β€M|cβa|n+1(n+1)!.
Exercises 8.5.6 Exercises
ΒΆ1. Determining Taylor polynomials from a function formula.
2. Determining Taylor polynomials from given derivative values.
3. Finding the Taylor series for a given rational function.
4. Finding the Taylor series for a given trigonometric function.
5. Finding the Taylor series for a given logarithmic function.
6.
In this exercise we investigation the Taylor series of polynomial functions.
Find the 3rd order Taylor polynomial centered at a=0 for f(x)=x3β2x2+3xβ1. Does your answer surprise you? Explain.
Without doing any additional computation, find the 4th, 12th, and 100th order Taylor polynomials (centered at a=0) for f(x)=x3β2x2+3xβ1. Why should you expect this?
Now suppose f(x) is a degree m polynomial. Completely describe the nth order Taylor polynomial (centered at a=0) for each n.
\(P_3(x) = -1 + 3x - \frac{4}{2!}x^2 + \frac{6}{3!}x^3 \text{,}\) which is the same polynomial as \(f(x)\text{.}\)
For \(n \ge 3\text{,}\) \(P_n(x) = f(x)\text{.}\)
For \(n \ge m\text{,}\) \(P_n(x) = f(x)\text{.}\)
-
Observe that \(f'(x) = 3x^2 - 4x + 3\text{,}\) \(f''(x) = 6x - 4\text{,}\) and \(f'''(x) = 6\text{,}\) so \(f'(0) = 3\text{,}\) \(f''(0) = -4\text{,}\) and \(f'''(0) = 6\text{.}\) In addition, \(f(0) = -1\) Thus, the 3rd order Taylor polynomial centered at \(a = 0\) is
\begin{equation*} P_3(x) = -1 + 3x - \frac{4}{2!}x^2 + \frac{6}{3!}x^3\text{.} \end{equation*}Note that this is the same polynomial as \(f(x)\text{,}\) which is not surprising since the degree 3 Taylor polynomial is the degree 3 polynomial that has the same value, slope, concavity, and third derivative at \(a = 0\) as the original.
Since \(f^{(4)}(x) = 0\) for all \(x\text{,}\) \(f^{(4)}(0) = 0\text{,}\) and similarly every higher derivative of \(f\) will be zero. Thus, the the 4th, 12th, and 100th order Taylor polynomials (centered at \(a = 0\)) are all identical to \(P_3(x)\) found in (a). for \(f(x) = x^3-2x^2+3x-1\text{.}\) We should expect this because all of the higher derivatives (than the third) of the original function are identically zero.
For \(n = 1, 2, \ldots, m-1\text{,}\) the degree \(n\) Taylor polyomial of \(f\) will be given by the usual formula. But for every \(n \ge m\text{,}\) \(P_n(x) = f(x)\text{.}\) That is, all of the Taylor polynomials of degree at least that of the original function are simply the function itself.
7.
The examples we have considered in this section have all been for Taylor polynomials and series centered at 0, but Taylor polynomials and series can be centered at any value of a. We look at examples of such Taylor polynomials in this exercise.
Let f(x)=sin(x). Find the Taylor polynomials up through order four of f centered at x=Ο2. Then find the Taylor series for f(x) centered at x=Ο2. Why should you have expected the result?
Let f(x)=ln(x). Find the Taylor polynomials up through order four of f centered at x=1. Then find the Taylor series for f(x) centered at x=1.
Use your result from (b) to determine which Taylor polynomial will approximate ln(2) to two decimal places. Explain in detail how you know you have the desired accuracy.
- \begin{align*} P_1(x) &= 1 + 0\left(x - \frac{\pi}{2} \right) = 1\\ P_2(x) &= 1 + 0\left(x - \frac{\pi}{2} \right) - \frac{1}{2!}\left(x - \frac{\pi}{2} \right)^2 = 1-\frac{1}{2!}\left(x - \frac{\pi}{2} \right)^2\\ P_3(x) &= 1 + 0\left(x - \frac{\pi}{2} \right) - \frac{1}{2!}\left(x - \frac{\pi}{2} \right)^2 + \frac{0}{3!}\left(x - \frac{\pi}{2} \right)^3\\ &= 1-\frac{1}{2!}\left(x - \frac{\pi}{2} \right)^2 = P_2(x)\\ P_4(x) &= 1 + 0\left(x - \frac{\pi}{2} \right) - \frac{1}{2!}\left(x - \frac{\pi}{2} \right)^2 + \frac{0}{3!}\left(x - \frac{\pi}{2} \right)^3 + \frac{1}{4!}\left(x - \frac{\pi}{2} \right)^4\\ &= 1 - \frac{1}{2!}\left(x - \frac{\pi}{2} \right)^2 + \frac{1}{4!}\left(x - \frac{\pi}{2} \right)^4\\ P(x) &= 1 - \frac{1}{2!}\left(x - \frac{\pi}{2} \right)^2 + \frac{1}{4!}\left(x - \frac{\pi}{2} \right)^4 - \frac{1}{6!}\left(x - \frac{\pi}{2} \right)^6 + \cdots \end{align*}
-
\begin{equation*} P_4(x) = 0 + 1(x-1) - \frac{1}{2!}(x-1)^2 + \frac{2}{3!}(x-1)^3 - \frac{6}{4!}(x-1)^4\text{.} \end{equation*}\begin{equation*} P(x) = 1(x-1) - \frac{1}{2}(x-1)^2 + \frac{1}{3}(x-1)^3 - \frac{1}{4}(x-1)^4 + \frac{1}{5}(x-1)^5 - \cdots \end{equation*}
\(P_{101}(1) \approx 0.698073\)
-
For \(f(x) = \sin(x)\text{,}\) \(f'(x) = \cos(x)\text{,}\) \(f''(x) = -\sin(x)\text{,}\) \(f'''(x) = -\cos(x)\text{,}\) and \(f^{(4)}(x) = \sin(x)\text{.}\) Thus, \(f\left(\frac{\pi}{2} \right) = 1\text{,}\) \(f'f\left(\frac{\pi}{2} \right) = 0\text{,}\) \(f''f\left(\frac{\pi}{2} \right) = -1\text{,}\) \(f'''f\left(\frac{\pi}{2} \right) = 0\text{,}\) and \(f^{(4)}f\left(\frac{\pi}{2} \right) = 1\text{.}\) It follows that the first four Taylor polynomials of \(f\) are
\begin{align*} P_1(x) &= 1 + 0\left(x - \frac{\pi}{2} \right) = 1\\ P_2(x) &= 1 + 0\left(x - \frac{\pi}{2} \right) - \frac{1}{2!}\left(x - \frac{\pi}{2} \right)^2 = 1-\frac{1}{2!}\left(x - \frac{\pi}{2} \right)^2\\ P_3(x) &= 1 + 0\left(x - \frac{\pi}{2} \right) - \frac{1}{2!}\left(x - \frac{\pi}{2} \right)^2 + \frac{0}{3!}\left(x - \frac{\pi}{2} \right)^3\\ &= 1-\frac{1}{2!}\left(x - \frac{\pi}{2} \right)^2 = P_2(x)\\ P_4(x) &= 1 + 0\left(x - \frac{\pi}{2} \right) - \frac{1}{2!}\left(x - \frac{\pi}{2} \right)^2 + \frac{0}{3!}\left(x - \frac{\pi}{2} \right)^3 + \frac{1}{4!}\left(x - \frac{\pi}{2} \right)^4\\ &= 1 - \frac{1}{2!}\left(x - \frac{\pi}{2} \right)^2 + \frac{1}{4!}\left(x - \frac{\pi}{2} \right)^4 \end{align*}From the pattern, the Taylor series for \(f(x)\) centered at \(x = \frac{\pi}{2}\) is
\begin{equation*} P(x) = 1 - \frac{1}{2!}\left(x - \frac{\pi}{2} \right)^2 + \frac{1}{4!}\left(x - \frac{\pi}{2} \right)^4 - \frac{1}{6!}\left(x - \frac{\pi}{2} \right)^6 + \cdots \end{equation*}which is expected because of the repeating patterns in the derivatives of the sine function evaluated at \(\frac{pi}{2}\text{.}\)
-
For \(f(x) = \ln(x)\text{,}\) \(f'(x) = x^{-1}\text{,}\) \(f''(x) = -x^{-2}\text{,}\) \(f'''(x) = 2x^{-3}\text{,}\) and \(f^{(4)}(x) = -6x^{-4}\text{.}\) It follows that \(f(1) = 0\text{,}\) \(f'(1) = 1\text{,}\) \(f''(1) = -1\text{,}\) \(f'''(1) = 2\text{,}\) and \(f^{(4)}(1) = -6\text{.}\) Thus, the fourth Taylor polynomial (in which we can see the polynomials of lower degree) is
\begin{equation*} P_4(x) = 0 + 1(x-1) - \frac{1}{2!}(x-1)^2 + \frac{2}{3!}(x-1)^3 - \frac{6}{4!}(x-1)^4\text{.} \end{equation*}Simplifying the coefficients and seeing the pattern, it follows that the Taylor series for \(f(x)\) centered at \(x = 1\) is
\begin{equation*} P(x) = 1(x-1) - \frac{1}{2}(x-1)^2 + \frac{1}{3}(x-1)^3 - \frac{1}{4}(x-1)^4 + \frac{1}{5}(x-1)^5 - \cdots \end{equation*} -
To estimate \(\ln(2)\) to two decimal places, we use the fact that \(\ln(x) = = 1(x-1) - \frac{1}{2}(x-1)^2 + \frac{1}{3}(x-1)^3 - \frac{1}{4}(x-1)^4 + \frac{1}{5}(x-1)^5 - \cdots\) for \(x\) near \(1\text{,}\) and thus
\begin{equation*} \ln(2) = P(2) = 1 - \frac{1}{2} + \frac{1}{3} - \frac{1}{4} + \cdots\text{.} \end{equation*}Because \(P(2)\) is an alternating series, we need to know when the next term in the series is less than \(0.01\text{,}\) which occurs when \(n = 101\text{.}\) Thus, if we compute \(P_{101}(1) = \sum_{k = 1}^{101} (-1)^{k+1} \frac{1}{k}\text{,}\) we get the desired estimate. We note particularly that \(P_{101}(1) \approx 0.698073\text{,}\) which is indeed an estimate of \(\ln(2)\) accurate to two decimal places.
8.
We can use known Taylor series to obtain other Taylor series, and we explore that idea in this exercise, as a preview of work in the following section.
Calculate the first four derivatives of sin(x2) and hence find the fourth order Taylor polynomial for sin(x2) centered at a=0.
-
Part (a) demonstrates the brute force approach to computing Taylor polynomials and series. Now we find an easier method that utilizes a known Taylor series. Recall that the Taylor series centered at 0 for f(x)=sin(x) is
ββk=0(β1)kx2k+1(2k+1)!.Substitute x2 for x in the Taylor series (8.5.7). Write out the first several terms and compare to your work in part (a). Explain why the substitution in this problem should give the Taylor series for sin(x2) centered at 0.
What should we expect the interval of convergence of the series for sin(x2) to be? Explain in detail.
\(P_4(x) = x^2 \text{.}\)
\(g(x) = P(x^2) = x^2 - \frac{1}{3!}x^6 + \frac{1}{5!}x^10 - \cdots \text{.}\)
All real numbers.
-
Let \(f(x) = \sin(x^2)\text{.}\) Then
\(\sin(x^2)\) \(f(0) = 0\) \(f'(x) = 2x\cos(x^2)\) \(f'(0) = 0\) \(f''(x) = -4x^2\sin(x^2) + 2\cos(x^2)\) \(f''(0) = 0\) \(f'''(x) = -8x^3\cos(x^2) - 12x\sin(x^2)\) \(f'''(0) = 0\) \(f^{(4)}(x) = 16x^4\sin(x^2) - 48x^2\cos(x^2) - 12\sin(x^2)\) \(f^{(4)}(0) = 0\) so the fourth order Taylor polynomial for \(f\) centered at \(a=0\) is
\begin{equation*} P_4(x) = x^2\text{.} \end{equation*} -
Substituting \(x^2\) for \(x\) in the Taylor series for \(\sin(x)\text{,}\) we find that since
\begin{equation*} P(x) = x - \frac{1}{3!}x^3 + \frac{1}{5!}x^5 - \cdots\text{,} \end{equation*}we have
\begin{equation*} g(x) = P(x^2) = x^2 - \frac{1}{3!}x^6 + \frac{1}{5!}x^10 - \cdots \end{equation*}which should be the Taylor series for \(\sin(x^2)\) since \(\sin(x) = x - \frac{1}{3!}x^3 + \frac{1}{5!}x^5 - \cdots\) for every value of \(x\text{.}\)
We expect the interval of convergence of the series for \(\sin(x^2)\) to be the same as the series for \(\sin(x)\text{,}\) since \(\sin(x) = x - \frac{1}{3!}x^3 + \frac{1}{5!}x^5 - \cdots\) for every value of \(x\text{.}\)
-
9.
Based on the examples we have seen, we might expect that the Taylor series for a function f always converges to the values f(x) on its interval of convergence. We explore that idea in more detail in this exercise. Let f(x)={eβ1/x2 if xβ 0,0 if x=0.
Show, using the definition of the derivative, that fβ²(0)=0.
It can be shown that f(n)(0)=0 for all nβ₯2. Assuming that this is true, find the Taylor series for f centered at 0.
What is the interval of convergence of the Taylor series centered at 0 for f? Explain. For which values of x the interval of convergence of the Taylor series does the Taylor series converge to f(x)?